{"id":"W2022902860","doi":"10.1145/1007568.1007621","title":"Incremental and effective data summarization for dynamic hierarchical clustering","year":2004,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":38,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Automatic summarization; Computer science; Cluster analysis; Data mining; Merge (version control); Hierarchical clustering; Data stream clustering; CURE data clustering algorithm; Fuzzy clustering; Information retrieval; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002310477,0.0008038736,0.0009177759,0.003168016,0.0007992564,0.00143616,0.002204511,0.0009464438,0.001156688],"category_scores_gemma":[0.01222314,0.0005312934,0.0007967071,0.003477201,0.0006753081,0.003083789,0.001996462,0.001036294,0.0006948797],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006824639,"about_ca_system_score_gemma":0.000900161,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00175082,"about_ca_topic_score_gemma":0.002321045,"domain_scores_codex":[0.9975495,0.0005997597,0.0002301419,0.000430108,0.001045586,0.0001449683],"domain_scores_gemma":[0.9940189,0.00202799,0.0006237486,0.001453679,0.001686009,0.0001896887],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005126261,0.0002205718,0.003381791,0.0004188544,0.0001405023,0.0002781022,0.0009131127,0.08704744,0.04435624,0.01627923,0.00952921,0.8369223],"study_design_scores_gemma":[0.00008010476,0.0003559575,0.002099648,0.00003503455,0.0001106431,0.0003720646,0.0003588395,0.9146551,0.04424725,0.02319075,0.01441576,0.00007879797],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01682502,0.0003926354,0.9804221,0.0001516135,0.00003953499,0.0001297864,0.0001995081,0.001411944,0.0004278184],"genre_scores_gemma":[0.1779446,0.0002704989,0.8186302,0.00009729466,0.00009768869,0.0002349763,0.001720711,0.0001972295,0.0008067947],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003168016,"threshold_uncertainty_score":0.01221913,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01748484195449467,"score_gpt":0.2759949414161231,"score_spread":0.2585100994616284,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}